[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-usability
Subject:    Re: Ctrl-Q vs. Ctrl-W vs. Alt-F4
From:       "Luciano Montanaro" <mikelima () cirulla ! net>
Date:       2009-01-15 15:46:33
Message-ID: b2289f80901150746s53d6a15au4209016b2904c718 () mail ! gmail ! com
[Download RAW message or body]

On Thu, Jan 15, 2009 at 3:22 PM, Maciej Pilichowski <bluedzins@wp.pl> wrote:
> On Thursday 15 January 2009 13:38:42 Luciano Montanaro wrote:
>
>> And which are the KDE applications for which loading time is so big
>> an issue? It may be interesting to fix them, if possible.
>
> I already gave examples: kdevelop, kate, kile. Non-kde: openoffice.

Kdevelop I concede may be slow to start; Kile takes maybe 3s here, and
I dont' find it too slow to be an annoyance to have to wait that much
for a restart. Maybe its startup time may be improved. I don't even
feel it needs a splash box anyway.

Kate startup is in the same ballpark the first time it runs, but it's
almost istantaneous on subsequent starts.

Finally, about OpenOffice: yes, it starts slowly. But we don't have a
say on that, and it implements the "Ctrl+W" quits the app behaviour
anyway.

If most applications outside KDE have this behaviour, the reasonable
expectation is for KDE apps to do the same...

>
>> > In result whole system runs slower because of usage of memory.
>> It should not. If an interactive application has no windows visible
>> and is not processing something on its own, it should not use CPU
>> at all, just some memory, which can be regarded as a sort of cache
>> for when the services of the application are needed again...
>> And if the memory pressure grows, it's likely to be paged out to
>> disk.
>
> Exactly. So there is cost.

Everything has a cost. Killing them has a cost, too.
But they are the likeliest thing to be paged out, since they are not in use.
Having to load them back has a higher cost, and keeping them in means
the linker has not to work resolving all those symbols again...

>
> (I beg to differ that launched app has no impact on CPU at all).
>
>> > You can't beat the physics -- computer resources are not
>> > unlimited.
>> But most computer nowadays have large amounts of memory that is
>> either unused or barely used as a disk cache.
>
> I didn't notice that because of such heavy load on my computer that I
> had to add switches before running any app to check if there is
> enough ram. Running amule + virtualbox + some computation stuff is a
> guarantee (at least for me) to have frozen computer. 2GB RAM.
>
> KDE has enough problems with running applications at the same time
> (like I am unable to scroll mail in KMail because Konqueror does
> something) already (2GHz CPU).
>
> However, this is going in wrong direction in general -- you would like
> (?) to just add more complexity to flawed design. Making whole
> architecture more complex, more error-prone will not hide the fact
> that the underlying source of those changes has flaws anyway.
>

I don't think it's wrong in general, and there are systems out there
ding exactly that, as someone else showed.

> Problem of relaunching application is just _one_ of the reasons why
> suggested UI should be rejected (and it is). Not the only one reason.
>
> Cheers,
> _______________________________________________
> kde-usability mailing list
> kde-usability@kde.org
> https://mail.kde.org/mailman/listinfo/kde-usability
>
>



-- 
Luciano Montanaro

Anyone who is capable of getting themselves made President should on
no account be allowed to do the job. -- Douglas Adams
_______________________________________________
kde-usability mailing list
kde-usability@kde.org
https://mail.kde.org/mailman/listinfo/kde-usability
[prev in list] [next in list] [prev in thread] [next in thread]